metadata
dataset_info:
features:
- name: image_url
dtype: string
- name: caption
dtype: string
splits:
- name: train
num_bytes: 602700071.1864096
num_examples: 2679713
download_size: 469085661
dataset_size: 602700071.1864096
configs:
- config_name: default
data_files:
- split: train
path: data/train-*